How air pollution altered the association of meteorological exposures and the incidence of dengue fever

Abstract

Background: Meteorological exposures are well-documented factors underlying the dengue pandemics, and air pollution was reported to have the potential to change the behaviors and health conditions of mosquitos. However, it remains unclear whether air pollution could modify the association of meteorological exposures and the incidence of dengue fever. Method: We matched the dengue surveillance data with the meteorological and air pollution data collected from monitoring sites from 2015 through 2019 in Guangzhou area. We developed generalized additive models with Poisson distribution to regress the daily counts of dengue against four meteorological exposures, while controlling for pollution and NDVI to evaluate the risk ratio (RR) of dengue for each unit increase in different exposures. The interaction terms of meteorological exposures and air pollution were then included to assess the modification effect of different pollution on the associations. Results: Daily dengue cases were nonlinearly associated with one-week cumulative temperature and precipitation, while not associated with humidity and wind speed. RRs were 1.07 (1.04, 1.11) and 0.95 (0.88, 1.03) for temperature below and above 27.1℃, 0.97 (0.96, 0.98) and 1.05 (1.01, 1.08) for precipitation below and above 20.3mm, respectively. For the modification effect, the RRs of low-temperature, wind speed on higher SO2 days and low-precipitation on both higher PM2.5 and SO2 days were greater compared to the low-pollution days with Pinteraction being 0.037, 0.030, 0.022 and 0.018. But the RRs of both high-temperature on higher SO2 days and high-precipitation on higher PM2.5 days were smaller with Pinteraction being 0.001 and 0.043. Conclusion: Air pollution could alter the meteorology-dengue associations. The impact of low-temperature, low-precipitation and wind speed on dengue occurrence tended to increase on days with high SO2 levels while the impact of high-temperature decreased. The impact of low-precipitation increased on high-PM2.5 days while the impact of high-precipitation decreased.